Introduction: In liver transplantation (LT), chronic kidney disease (CKD) prediction and management improves morbidity and mortality, both of which are challenging and no Indian study has explored this field of research. We aimed to develop a predictor model to detect KDIGO CKD 3a or higher with pre-operative risk factors.
Methods: In this study of liver donor LT a derivation cohort (retrospective arm, n = 196) and a validation cohort (prospective arm, n =46) was studied during 2017 to 2020. Outcome assessed was stage CKD 3a or higher.
Results: The incidence of CKD 3a or higher in derivation and validation cohort was 50/196(25.5%) and 9/46(19.5%) respectively with a follow-up of 3 years.  There was statistically significant difference between derivation and validation cohort for following pre-operative risk factors: pre-transplant eGFR (estimated glomerular filtration rate), pre-operative platelet count and pre-operative acute kidney injury (AKI). The predictor equation for development of CKD 3a after LDLT was R = -0.2691 + 1.1193 x diabetes – 0.0213 * pretransplant eGFR + 2.0542*pre-AKI, with optimal cut-off = 0.03176. A C-statistics of 0.7237 was obtained in validation cohort, with sensitivity and negative predictive value of 77% and 92.31% respectively
Conclusion: We report the first Indian study which predicts development of high risk CKD in liver transplant population. As the risk factors for CKD development in Indian context is different from western data, so this report will help in improving LDLT outcomes in emerging nations like India.